提出一种基于粗糙模糊集的层次聚类算法。在该算法中,通过事务间的相似性关系生成用户事务模糊矩阵,运用基于模糊矩阵的粗糙上近似提取初始类,使用约束相似性的上近似概念形成后续类,其中的一个相对的相似性条件被用作合并准则。该算法可以帮助Web挖掘者描述潜在的新的Web用户组特性。
This paper presents an agglomerative hierarchical clustering algorithm based on rough-fuzzy set. In this algorithm, a user session fuzzy matrix is made through session's similarity and initial clusters are formed using rough upper approximation based on the fuzzy matrix. Subsequent clusters are formed using the concept of constrained-similarity upper approximation wherein a condition of relative similarity is used as a merging criteri- on. This algorithm can help Web miners to identify potential groups of users.